U.S. patent number 8,107,773 [Application Number 11/861,543] was granted by the patent office on 2012-01-31 for video signal processing apparatus and video signal processing method.
This patent grant is currently assigned to Kabushiki Kaisha Toshiba. Invention is credited to Shin Arai, Yasuhiro Hori, Koichi Sato.
United States Patent |
8,107,773 |
Hori , et al. |
January 31, 2012 |
Video signal processing apparatus and video signal processing
method
Abstract
A video-signal processor includes: a first correlation detector
obtaining a correlation value of every symmetric pair of pixels
positioned in point-symmetry around an interpolation pixel on an
interpolation line interpolated between two real lines; a second
correlation detector obtaining a correlation value of every pair of
pixels for the symmetric pair of pixels and parallel pair(s) of
pixels positioned in a direction parallel to a direction of the
symmetric pair of pixels; a correlation-direction deciding unit
deciding one of a correlation direction based on the correlation
value from the first correlation detector and that based on the
correlation value from the second correlation detector as a
correlation direction which is a direction of the pair of pixels
used for generating the interpolation pixel; and an interpolation
pixel generator generating a pixel signal of the interpolation
pixel using a pair of pixels based on decision result of the
correlation-direction deciding unit.
Inventors: |
Hori; Yasuhiro (Tokyo,
JP), Arai; Shin (Tokyo, JP), Sato;
Koichi (Kanagawa, JP) |
Assignee: |
Kabushiki Kaisha Toshiba
(Tokyo, JP)
|
Family
ID: |
39261283 |
Appl.
No.: |
11/861,543 |
Filed: |
September 26, 2007 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20080080790 A1 |
Apr 3, 2008 |
|
Foreign Application Priority Data
|
|
|
|
|
Sep 27, 2006 [JP] |
|
|
2006-263197 |
|
Current U.S.
Class: |
382/300 |
Current CPC
Class: |
H04N
7/012 (20130101); G09G 3/20 (20130101); H04N
7/0135 (20130101); G09G 2310/0229 (20130101) |
Current International
Class: |
G06K
9/32 (20060101) |
Field of
Search: |
;382/300 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
1574951 |
|
Feb 2005 |
|
CN |
|
2003-230109 |
|
Aug 2003 |
|
JP |
|
2005-293361 |
|
Oct 2005 |
|
JP |
|
Other References
Chinese Office Action for 2007101691774 mailed on Jan. 15, 2010.
cited by other .
Japanese Office Action for Japanese Patent Application Serial No.
2006-263197 mailed on Jul. 12, 2011. cited by other.
|
Primary Examiner: Tucker; Wesley
Attorney, Agent or Firm: Turocy & Watson, LLP
Claims
What is claimed is:
1. A video signal processing apparatus comprising, a first
correlation detecting unit obtaining a correlation value of every
symmetric pair of pixels positioned in a point-symmetry, center of
which is an interpolation pixel on an interpolation line
interpolated between two real lines, the symmetric pair of pixels
being positioned on the two real lines of an input video signal
respectively; a second correlation detecting unit obtaining a
correlation value of every pair of pixels for the symmetric pair of
pixels and one or more parallel pairs of pixels positioned in a
direction which is parallel to a direction of the symmetric pair of
pixels, the parallel pair of pixels being positioned on the two
real lines of the input video signal respectively; a correlation
direction deciding unit deciding any one of a correlation direction
based on the correlation value from the first correlation detecting
unit and a correlation direction based on the correlation value
from the second correlation detecting unit as a correlation
direction which is a direction of a pair of pixels used for
generating the interpolation pixel; an interpolation pixel
generating unit generating a pixel signal of the interpolation
pixel by using the pair of pixels based on a decision result of the
correlation direction deciding unit; and a horizontal high
frequency component detecting unit detecting a horizontal high
frequency component of a pair of pixels positioned on the two real
lines of the input video signal respectively, wherein when it is
decided based on a detection result by the horizontal high
frequency component detecting unit that the symmetric pair of
pixels are pixels in a horizontal high frequency component area,
the correlation direction deciding unit designates the correlation
direction based on the correlation value from the second
correlation detecting unit to be a decision result.
2. The video signal processing apparatus according to claim 1,
wherein the number of the symmetric pairs of pixels, the
correlation value of which is obtained by the second correlation
detecting unit, is less than the number of the symmetric pairs of
pixels, the correlation value of which is obtained by the first
correlation detecting unit.
3. The video signal processing apparatus according to claim 1,
wherein when it is decided based on a detection result by the
horizontal high frequency component detecting unit that the
symmetric pair of pixels are pixels in a horizontal high frequency
component area, for the correlation value of every symmetric pair
of pixels from the second correlation detecting unit, the
correlation direction deciding unit calculates the sum of the
correlation value of the symmetric pair of pixels and the
correlation values of the one or more parallel pairs of pixels
positioned in a direction which is parallel to a direction of the
symmetric pair of pixels, and designates the correlation direction
indicating the symmetric pair of pixels, the sum of the correlation
values of which is minimum, to be a decision result.
4. A video signal processing apparatus comprising, a first
correlation detecting unit obtaining a correlation value of every
symmetric pair of pixels positioned in a point-symmetry, center of
which is an interpolation pixel on an interpolation line
interpolated between two real lines, the symmetric pair of pixels
being positioned on the two real lines of an input video signal
respectively; a second correlation detecting unit obtaining a
correlation value of every pair of pixels for the symmetric pair of
pixels and one or more parallel pairs of pixels positioned in a
direction which is parallel to a direction of the symmetric pair of
pixels, the parallel pair of pixels being positioned on the two
real lines of the input video signal respectively; a correlation
direction deciding unit deciding any one of a correlation direction
based on the correlation value from the first correlation detecting
unit and a correlation direction based on the correlation value
from the second correlation detecting unit as a correlation
direction which is a direction of a pair of pixels used for
generating the interpolation pixel; an interpolation pixel
generating unit generating a pixel signal of the interpolation
pixel by using the pair of pixels based on a decision result of the
correlation direction deciding unit; a temporary interpolation
pixel generating unit generating a temporary interpolation pixel
signal by using the pair of pixels of the correlation direction
based on the correlation value from the first correlation detecting
unit; and a peripheral pixel selecting unit outputting the pixel
signal of the pair of pixels around the interpolation pixel, the
pair of pixels being positioned on the two real lines of the input
video signal respectively, wherein when a value of the temporary
interpolation pixel signal is a maximum value or a minimum value of
the pixel signals from the peripheral pixel selecting unit, the
correlation direction deciding unit designates the correlation
direction based on the correlation value from the second
correlation detecting unit to be a decision result.
5. The video signal processing apparatus according to claim 4,
wherein the number of the symmetric pairs of pixels, the
correlation value of which is obtained by the second correlation
detecting unit, is less than the number of the symmetric pairs of
pixels, the correlation value of which is obtained by the first
correlation detecting unit.
6. The video signal processing apparatus according to claim 4,
wherein when a value of the temporary interpolation pixel signal is
a maximum value or a minimum value of the pixel signals from the
peripheral pixel selecting unit, for the correlation value of every
symmetric pair of pixels from the second correlation detecting
unit, the correlation direction deciding unit calculates the sum of
the correlation value of the symmetric pair of pixels and the
correlation values of the one or more parallel pairs of pixels
positioned in a direction which is parallel to a direction of the
symmetric pair of pixels, and designates the correlation direction
indicating the symmetric pair of pixels, the sum of the correlation
values of which is minimum, to be a decision result.
7. A video signal processing apparatus comprising, a first
correlation detecting unit obtaining a correlation value of every
symmetric pair of pixels positioned in a point-symmetry, center of
which is an interpolation pixel on an interpolation line
interpolated between two real lines, the symmetric pair of pixels
being positioned on the two real lines of an input video signal
respectively; a second correlation detecting unit obtaining a
correlation value of every pair of pixels for the symmetric pair of
pixels and one or more parallel pairs of pixels positioned in a
direction which is parallel to a direction of the symmetric pair of
pixels, the parallel pair of pixels being positioned on the two
real lines of the input video signal respectively; a correlation
direction deciding unit deciding any one of a correlation direction
based on the correlation value from the first correlation detecting
unit and a correlation direction based on the correlation value
from the second correlation detecting unit as a correlation
direction which is a direction of a pair of pixels used for
generating the interpolation pixel; and an interpolation pixel
generating unit generating a pixel signal of the interpolation
pixel by using the pair of pixels based on a decision result of the
correlation direction deciding unit, wherein when a minimum value
of the correlation values from the first correlation detecting unit
is no less than a first threshold, the correlation direction
deciding unit causes the correlation direction based on the
correlation values from the first correlation detecting unit to be
invalid, and when a minimum value of the correlation values from
the second correlation detecting unit is no less than a second
threshold which is different from the first threshold, the
correlation direction deciding unit causes the correlation
direction based on the correlation values from the second
correlation detecting unit to be invalid.
8. The video signal processing apparatus according to claim 7,
wherein the number of the symmetric pairs of pixels, the
correlation value of which is obtained by the second correlation
detecting unit, is less than the number of the symmetric pairs of
pixels, the correlation value of which is obtained by the first
correlation detecting unit.
9. The video signal processing apparatus according to claim 7,
wherein when a minimum value of the correlation values from the
first correlation detecting unit is no less than the first
threshold, and a minimum value of the correlation values from the
second correlation detecting unit is no less than the second
threshold which is different from the first threshold, the
correlation direction deciding unit causes the correlation
direction indicating the symmetric pair of pixels in a horizontally
same position as that of the interpolation pixel to be a decision
result.
Description
CROSS REFERENCE TO RELATED APPLICATION
This application is based upon and claims the benefit of priority
from the prior Japanese Patent Application No. 2006-263197 filed on
Sep. 27, 2006; the entire contents of which are incorporated herein
by reference.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a video signal processing
apparatus and a video signal processing method which are suitable
for a scanning line interpolation circuit, etc. configured to
convert scanning lines from a video signal of an interlaced mode to
a video signal of a progressive mode.
2. Description of the Related Art
A plasma display apparatus, a liquid crystal display apparatus,
etc. configure a matrix display apparatus in which pixels are
placed in a matrix state corresponding to intersecting points of a
plurality of scanning lines and data lines. In the matrix display
apparatus, video information per pixel is sequentially displayed
for every scanning line so that an image is displayed.
Some of such matrix display apparatuses have a display screen whose
resolution is higher than the resolution of an input video signal.
In such a case, a scanning line interpolation circuit may be
adopted which improves the image quality by setting a interpolation
line between scanning lines based on inputted video signals, and
generating interpolation pixels on the interpolation line.
Some scanning line interpolation circuit may generate the
interpolation pixel according to an average value of pixels of
upper and under lines of the interpolation line. However, this
interpolation method may generate such an image that a diagonal
line looks as jagging, what we say, jaggy is conspicuous in some
patterns.
Thus, Japanese Patent Laid-open No. 2003-230109 proposes a line
scanning line interpolation circuit which averages
highly-correlative diagonal direction pixel values to prevent the
resolution degradation of diagonal edges. In this proposal, in the
respective pixels of upper and under lines of the interpolation
line, a pair of pixels are used for the interpolation, which is
positioned at a point-symmetric point whose center is an
interpolation pixel of the interpolation line. The correlation is
detected for each of such pair of pixels, and a direction of a pair
of pixels whose correlation is highest is decided to be a direction
of a pixel (correlation direction) whose correlation is high with
respect to the interpolation pixel. Next, the interpolation pixel
is generated by using two pixels positioned in the correlation
direction.
However, in a pattern part which is relatively slender, diagonal,
and line-like, etc., the correlation direction may not be correctly
specified. For example, it is assumed that a pair of pixels for
specifying the correlation direction is positioned at background
parts of both sides crossing over the line-like pattern. In such a
case, if the brightness of the background parts is approximately
the same, it may be considered that the correlation direction
obtained for the interpolation pixel in the line-like pattern is
specified as not a direction of the line but a direction of the
pair of pixels positioned at the background. If so, the
interpolation pixel in the line-like pattern is interpolated by
using two pixels of the background, then breaks, unsharpness, jaggy
may be induced in the line-like pattern.
SUMMARY OF THE INVENTION
A video signal processing apparatus according to an aspect of the
present invention includes: a first correlation detecting unit
obtaining a correlation value of every symmetric pair of pixels
positioned in a point-symmetry center of which is an interpolation
pixel on an interpolation line interpolated between two real lines,
the symmetric pair of pixels being positioned on the two real lines
of an input video signal respectively; a second correlation
detecting unit obtaining a correlation value of every pair of
pixels for the symmetric pair of pixels and one or more parallel
pairs of pixels positioned in a direction which is parallel to a
direction of the symmetric pair of pixels, the parallel pair of
pixels being positioned on the two real lines of the input video
signal respectively; a correlation direction deciding unit deciding
any one of a correlation direction based on the correlation value
from the first correlation detecting unit and a correlation
direction based on the correlation value from the second
correlation detecting unit as a correlation direction which is a
direction of a pair of pixels used for generating the interpolation
pixel; and an interpolation pixel generating unit generating a
pixel signal of the interpolation pixel by using the pair of pixels
based on a decision result of the correlation direction deciding
unit.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram illustrating a video signal processing
apparatus according to a first embodiment of the present
invention;
FIG. 2 is an explanatory diagram describing a first correlation
detecting unit 20;
FIG. 3 is an explanatory diagram describing a second correlation
detecting unit 30;
FIG. 4 is an explanatory diagram describing the second correlation
detecting unit 30;
FIG. 5 is an explanatory diagram describing the second correlation
detecting unit 30;
FIG. 6 is a block diagram illustrating a second embodiment of the
present invention;
FIG. 7 is an explanatory diagram describing operations of the
second embodiment;
FIG. 8 is a block diagram illustrating a third embodiment of the
present invention;
FIG. 9 is a block diagram illustrating a fourth embodiment of the
present invention; and
FIG. 10 is a flowchart illustrating operations of the fourth
embodiment.
DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTION
Embodiments of the present invention will be described in detail
below referring to the drawings.
First Embodiment
FIG. 1 is a block diagram illustrating a video signal processing
apparatus according to a first embodiment of the present invention.
The present embodiment is an example applied to a scanning line
interpolation circuit.
In FIG. 1, a video signal is inputted to an input terminal 11. The
input video signal is provided to a line memory 12. The line memory
12 memorizes a line (horizontal period) of inputted video signal.
Input and output of the line memory 12 are video signals before and
after one line. The input video signal from the input terminal 11
and the video signal from the line memory 12 are delivered to first
and second correlation detecting units 20 and 30, and a horizontal
high frequency detecting circuit 35.
The first correlation detecting unit 20 detects a pair of pixels
whose correlation with an interpolation pixel is high to select a
pair of pixels used for generating a pixel to be interpolated
(interpolation pixel) on an interpolation line. That is, the first
correlation detecting unit 20 obtains a correlation of each pair of
pixels which are positioned at a point-symmetric point whose center
is an interpolation pixel in the respective pixels on upper and
under lines (real line) of the interpolation line.
A symmetric pixel selecting circuit 21 includes a memory which is
not illustrated, holds pixel signals whose position is
point-symmetric with the interpolation pixel in inputted video
signals (pixel signal) of the respective pixels of two lines, and
outputs to a correlation value calculation circuit 22.
FIG. 2 is an explanatory diagram describing the first correlation
detecting unit 20. In FIG. 2, horizontal lines A and B represent
two lines which are adjacent to each other in an image. A mark;
represents a pixel position. A horizontal line P represents an
interpolation line, and a pixel P0 on the interpolation line P
represents an interpolation pixel to be generated.
The first correlation detecting unit 20 sets a prescribed
correlation direction candidate whose center is the interpolation
pixel P0. FIG. 2 illustrates such an example that seven directions
(L1 to L7) are set as the correlation direction candidates by using
pixels of seven pixels.times.two lines. The symmetric pixel
selecting circuit 21 of the first correlation detecting unit 20
holds the respective pixels signal in a correlation direction
detecting range, selects point-symmetric pairs of pixels (A1, B7),
(A2, B6), - - - , (A7, B1) whose center is the interpolation pixel
P0 respectively, and outputs pixel signals of each pair of pixels
to the correlation value calculation circuit 22. The correlation
value calculation circuit 22 calculates a correlation value of each
pair of pixels respectively. The correlation value calculation
circuit 22 outputs the obtained correlation value to a correlation
direction deciding unit 40.
Conventionally, a correlation direction candidate is decided to be
a correlation direction, which corresponds to a pair of pixels
providing the highest correlation in respective correlation values
obtained by the first correlation detecting unit 20, and an
interpolation pixel is generated by using two pixels positioned in
the correlation direction.
On the other hand, in the present embodiment, the correlation
direction deciding unit 40 utilizes not only output of the first
correlation detecting unit 20 but also output of the second
correlation detecting unit 30. The second correlation detecting
unit 30 detects a correlation direction by using a pair of pixels
positioned in a direction which is parallel with a direction of
each pair of pixels used by the first correlation detecting unit
20.
FIG. 3 to FIG. 5 are explanatory diagrams describing the second
correlation detecting unit 30. In FIG. 3 to FIG. 5, horizontal
lines A, B, P and the respective pixels represent the same lines
and pixels as those of FIG. 2.
A symmetric and parallel pixel selecting circuit 31 of the second
correlation detecting unit 30 selects a pair of pixels at a
position which is point-symmetric to an interpolation pixel, as a
pair of pixels for obtaining a correlation direction, like the
symmetric pixel selecting circuit 21. Further, the symmetric and
parallel pixel selecting circuit 31 selects one or more pairs of
pixels positioned in a direction which is parallel to a direction
of the selected pair of pixels at a point-symmetric position. That
is, the symmetric and parallel pixel selecting circuit 31
designates two pixels which are shifted by the same distance and in
the same direction respectively with respect to the respective
pixels at a point which is point-symmetric to an interpolation
pixel to be a pair of pixels used for detecting a correlation
direction.
In such a case, the second correlation detecting unit 30 sets
correlation direction candidates whose number is smaller than the
number of the candidates set by the first correlation detecting
unit 20 as correlation direction candidates to be set. FIG. 3 to
FIG. 5 illustrate such examples that one of three correlation
direction candidates is detected by using pixels of 5
pixels.times.two lines.
In examples of FIG. 3 to FIG. 5, the symmetric and parallel pixel
selecting circuit 31 sets three directions as correlation direction
candidates, and selects pairs of pixels used for deciding
correlations of three set directions. That is, as illustrated in
FIG. 3, the symmetric and parallel pixel selecting circuit 31
selects the symmetric pair of pixels (A4, B4) as a pair of pixels
for detecting a vertical correlation direction candidate, sets two
directions M4 and N4 which are parallel to a direction L4 of this
pair of pixels, and selects pairs of pixels (A3, B3), (A5, B5) in
these directions as parallel pairs of pixels. As illustrated in
FIG. 4, the symmetric and parallel pixel selecting circuit 31
selects the symmetric pair of pixels (A3, B5) as a pair of pixels
for detecting a left and diagonal correlation direction candidate,
sets two directions M3 and N3 which are parallel to a direction L3
of this pair of pixels, and selects pairs of pixels (A2, B4), (A4,
B6) in these directions as parallel pairs of pixels. Furthermore,
as illustrated in FIG. 5, the symmetric and parallel pixel
selecting circuit 31 selects the symmetric pair of pixels (A5, B3)
as a pair of pixels for detecting a right and diagonal correlation
direction candidate, sets two directions M5 and N5 which are
parallel to a direction L5 of this pair of pixels, and selects
pairs of pixels (A4, B2), (A6, B4) in these directions as parallel
pairs of pixels.
While in examples of FIG. 3 to FIG. 5, the symmetric and parallel
pixel selecting circuit 31 sets three pairs of pixels as
correlation direction candidates to be detected, it is apparent
that four or more pairs of pixels may be set. The symmetric and
parallel pixel selecting circuit 31 outputs pixel signals of
respective selected pair of pixels to a correlation value
calculation circuit 32. The correlation value calculation circuit
32 calculates a correlation value of each pair of pixels
respectively. The correlation value calculation circuit 32 outputs
the obtained correlation value to the correlation direction
deciding unit 40.
The correlation direction deciding unit 40 includes a memory unit
holding a correlation value of each pair of pixels from the
correlation value calculation circuits 22 and 32. The correlation
direction deciding unit 40 compares with each other the correlation
values from the correlation value calculation circuit 22, and
decides that a correlation direction candidate corresponding to a
pair of pixels whose correlation value is minimum is a correlation
direction obtained by the first correlation detecting unit 20. The
correlation direction deciding unit 40 obtains the sum of
correlation values of three pairs of pixels based on the same
correlation direction candidate in correlation values from the
correlation value calculation circuit 32 for every correlation
direction candidate, and decides that a correlation direction
candidate whose sum of correlation values is minimum is a
correlation direction obtained by the second correlation detecting
unit 30. The correlation direction deciding unit 40 outputs such a
decision result to a correlation direction determining unit 41 that
either one of a correlation direction obtained by the first
correlation detecting unit 20 and a correlation direction obtained
by the second correlation detecting unit 30 is designated to be a
correlation direction.
Because the correlation direction obtained by the first correlation
detecting unit 20 is selected from many correlation direction
candidates, it is highly accurate. On the other hand, the
correlation direction obtained by the second correlation detecting
unit 30 is obtained by such a calculation that a plurality of pairs
of pixels are used for one correlation direction candidate in few
correlation direction candidates, so that wrong decisions are few.
That is, the correlation direction deciding unit 40 according to
the present embodiment can select whether a correlation direction
is obtained in high accuracy, or without wrong decisions.
For example, the correlation direction deciding unit 40 can select
and output a correlation direction obtained by the first
correlation detecting unit 20 and a correlation direction obtained
by the second correlation detecting unit 30 based on a detection
result by the horizontal high frequency detecting circuit 35.
Video signals before and after an interpolation line are inputted
to the horizontal high frequency detecting circuit 35. The
horizontal high frequency detecting circuit 35 detects a horizontal
high frequency component of lines before and after an interpolation
line. The horizontal high frequency detecting circuit 35 obtains
the horizontal high frequency components for pixels selected for
detecting a correlation direction in the first and second
correlation detecting units 20 and 30. The horizontal high
frequency detecting circuit 35 compares each obtained horizontal
high frequency component with a prescribed threshold, and decides
whether or not a pixel of a correlation direction candidate is
included in a horizontal high frequency component area according to
the number of horizontal high frequency components which are no
less than the prescribed threshold. The horizontal high frequency
detecting circuit 35 delivers the decision result to the
correlation direction deciding unit 40.
Meanwhile, the horizontal high frequency component area is an area
in which a horizontal frequency is high, for example, in such a
pattern part of thickly growing leaves, a pattern part of slender
branches, etc., if only a correlation value calculation by using a
symmetric pair of pixels is used, wrong decision for a correlation
direction is easily induced.
When the decision result of the horizontal high frequency detecting
circuit 35 is provided, and the pair of pixels of a correlation
direction obtained by the first correlation detecting unit 20 is
included in the horizontal high frequency component area, the
correlation direction deciding unit 40 selects a correlation
direction obtained by the second correlation detecting unit 30 and
outputs to the correlation direction determining unit 41. When the
pair of pixels of a correlation direction obtained by the first
correlation detecting unit 20 is not included in the horizontal
high frequency component area, the correlation direction deciding
unit 40 selects a correlation direction obtained by the first
correlation detecting unit 20 and outputs to the correlation
direction determining unit 41.
The correlation direction determining unit 41 determines a pair of
pixels used for generating an interpolation pixel based on a
decision result of a correlation direction by the correlation
direction deciding unit 40. That is, the correlation direction
determining unit 41 determines a pair of pixels on a correlation
direction as a pair of pixels used for generating an interpolation
pixel. The correlation direction determining unit 41 indicates the
pair of pixels used for generating an interpolation pixel to an
interpolation pixel generating circuit 42. The interpolation pixel
generating circuit 42 includes a memory unit which is not
illustrated and memorizes video signals of lines before and after
an interpolation line. The interpolation pixel generating circuit
42 generates an interpolation pixel with a prescribed calculation,
e.g. an average calculation, using the indicated pair of pixels.
The interpolation pixel generating circuit 42 outputs the generated
interpolation pixel as an interpolation signal to an output
terminal 43.
Next, operations of the embodiment configured as above will be
described.
An input video signal inputted through the input terminal 11 is
delayed by one horizontal period in the line memory 12. A video
signal from the input terminal 11 and output of the line memory 12
are delivered to the first and second correlation detecting units
20 and 30.
The symmetric pixel selecting circuit 21 of the first correlation
detecting unit 20 selects a symmetric pair of pixels of an
interpolation pixel to output signals of the pixels to the
correlation value calculation circuit 22. The first correlation
detecting unit 20 selects, for example, pairs of pixels of the
seven correlation direction candidates of FIG. 2. The correlation
value calculation circuit 22 obtains a correlation value for every
pair of pixels of each correlation direction candidate, and outputs
a calculated result to the correlation direction deciding unit 40.
The correlation direction deciding unit 40 designates a correlation
direction candidate which provides a minimum correlation value of
correlation values from the correlation value calculation circuit
22 to be a correlation direction by the first correlation detecting
unit 20.
The symmetric and parallel pixel selecting circuit 31 of the second
correlation detecting unit 30 selects a symmetric pair of pixels of
an interpolation pixel, and also selects two pixels which are
parted by the same distance and in the same direction from the
respective pixels of the selected pair of pixels as a parallel pair
of pixels. For example, as illustrated in FIG. 3 to FIG. 5, the
symmetric and parallel pixel selecting circuit 31 selects three
pairs of pixels on three correlation direction candidates. The
symmetric and parallel pixel selecting circuit 31 outputs pixel
signals of each selected pair of pixels to the correlation value
calculation circuit 32. The correlation value calculation circuit
32 obtains a correlation value for every pair of pixels of each
correlation direction candidate, and outputs the calculated result
to the correlation direction deciding unit 40. The correlation
direction deciding unit 40 adds the correlation values from the
correlation value calculation circuit 32 for every correlation
direction candidate, and designates a correlation direction
candidate whose sum is minimum to be a correlation direction by the
second correlation detecting unit 30.
On the other hand, video signals of lines before and after an
interpolation line are also delivered to the horizontal high
frequency detecting circuit 35. The horizontal high frequency
detecting circuit 35 detects a horizontal high frequency component
for each pair of pixels selected by the first and second
correlation detecting units 20 and 30 for detecting a correlation
direction. The horizontal high frequency detecting circuit 35
compares each obtained horizontal high frequency component with a
prescribed threshold, and decides whether or not a pixel of a
correlation direction candidate is included in a horizontal high
frequency component area according to the number of horizontal high
frequency components which are no less than the prescribed
threshold. The horizontal high frequency detecting circuit 35
delivers the decision result to the correlation direction deciding
unit 40.
The correlation direction deciding unit 40 decides whether or not a
pair of pixels of a correlation direction obtained by the first
correlation detecting unit 20 is included in a horizontal high
frequency component area based on the decision result of the
horizontal high frequency detecting circuit 35. When the pair of
pixels of a correlation direction obtained by the first correlation
detecting unit 20 is included in a horizontal high frequency
component area, the correlation direction deciding unit 40 selects
a correlation direction by the second correlation detecting unit
30, and if not, selects a correlation direction by the first
correlation detecting unit 20.
As described above, when, for example, an interpolation pixel is a
pixel in a pattern of slender branches, if only a detection result
by the first correlation detecting unit 20 is used for deciding a
correlation direction, a correlation direction candidate from pairs
of pixels of a pattern part of a background may be selected as a
correlation direction. Even in such a case, the pattern part is
decided to be a horizontal high frequency component area, so that
the correlation direction deciding unit 40 selects a correlation
direction of the second correlation detecting unit 30. That is, it
is possible to prevent an interpolation pixel from being generated
by a pair of pixels of a background part.
The second correlation detecting unit 30 detects a correlation of
each correlation direction candidate with pixels on a line
including an interpolation pixel and pixels on lines which are
parallel to the line including the interpolation pixel. Thus, when
an interpolation pixel is a pixel in a branch pattern, such a
possibility is low that the sum of correlation values of
correlation direction candidates in a direction which is widely
shifted from a direction in which a branch grows becomes small, and
such a possibility is high that the sum of correlation values of
correlation direction candidates in a direction which is close to a
direction in which a branch of a pattern grows becomes small. That
is, if the correlation direction deciding unit 40 selects a
correlation direction by the second correlation detecting unit 30
by using a decision result of the horizontal high frequency
detecting circuit 35, it is possible to detect a correlation
direction with few wrong decisions.
The correlation direction determining unit 41 determines a pair of
pixels used for generating an interpolation pixel based on a
decision result of a correlation direction by the correlation
direction deciding unit 40. The correlation direction determining
unit 41 indicates the pair of pixels used for generating an
interpolation pixel to the interpolation pixel generating circuit
42. The interpolation pixel generating circuit 42 generates an
interpolation pixel with a prescribed calculation, e.g. an average
calculation, using the indicated a pair of pixels. Meanwhile, even
if a correlation direction by the second correlation detecting unit
30 is selected, a pair of pixels, symmetric pair of pixels, is
indicated to the interpolation pixel generating circuit 42 based on
the correlation direction. The interpolation pixel generating
circuit 42 outputs the generated interpolation pixel as an
interpolation signal.
As described above, in the present embodiment, the first
correlation detecting unit 20 obtains a correlation direction in
high accuracy which is a direction of a pair of pixels used for
generating an interpolation pixel, and the second correlation
detecting unit 30 obtains a correlation direction without wrong
decisions. By deciding a horizontal high frequency component area
with the horizontal high frequency detecting circuit 35, one of
correlation directions by the first and second correlation
detecting units 20 and 30 is selected, so that it is possible to
prevent wrong decisions and generate an interpolation pixel in high
accuracy. Thereby, it is possible to interpolate by using pixels in
the same pattern, and even in a slender line-like pattern part,
etc., it is possible to interpolate without inducing breaks,
unsharpness, jaggy, etc.
Second Embodiment
FIG. 6 is a block diagram illustrating a second embodiment of the
present invention. In FIG. 6, the same code is attached to the same
component as that of FIG. 1, and the description will be
omitted.
The present embodiment also adopts the first and second correlation
detecting units 20 and 30. In the present embodiment, it is
determined to use which one of outputs from the first and second
correlation detecting units 20 and 30 by deciding the reliability
of a generated interpolation signal.
The present embodiment adopts a peripheral pixel selecting circuit
55 instead of the horizontal high frequency detecting circuit 35,
and also adopts a correlation direction deciding unit 50 instead of
the correlation direction deciding unit 40. The peripheral pixel
selecting circuit 55 selects a pixel around an interpolation pixel
in pixels of lines before and after an interpolation line to output
a signal of the pixel to the correlation direction deciding unit
50.
The correlation direction deciding unit 50 is provided with outputs
of the first and second correlation detecting units 20 and 30. The
correlation direction deciding unit 50, first, obtains a
correlation direction by using output of the first detecting unit
20. The correlation direction deciding unit 50 outputs the decision
result of a correlation direction by the first correlation
detecting unit 20 to the correlation direction determining unit 41.
The correlation direction determining unit 41 indicates a pair of
pixels based on a correlation direction by the first correlation
detecting unit 20 to the interpolation pixel generating circuit 42
based on the decision result of the correlation direction deciding
unit 50. The interpolation pixel generating circuit 42 obtains a
pixel signal with, e.g. an average calculation by using the
indicated pair of pixels.
In the present embodiment, the interpolation pixel generating
circuit 42 outputs the thus obtained pixel signal as a temporary
interpolation pixel signal to the correlation direction deciding
unit 50. The correlation direction deciding unit 50 includes a
maximum value and minimum value deciding unit 51. The maximum value
and minimum value deciding unit 51 is delivered with a pixel signal
from the peripheral pixel selecting circuit 55 and a temporary
interpolation pixel signal. The maximum value and minimum value
deciding unit 51 compares the pixel signal around an interpolation
pixel with the temporary interpolation pixel signal to decide
whether or not the temporary interpolation pixel signal is a
maximum value or a minimum value of a pixel signal of the
peripheral pixel.
When there is not a wrong decision in the decision result of a
correlation direction, it is highly possible that a value of the
temporary interpolation pixel signal is relatively close to that of
a pixel signal of the peripheral pixel. On the contrary, when a
value of the temporary interpolation pixel signal is relatively
apart from that of a pixel signal of the peripheral pixel, it is
highly possible that there is a wrong decision in the decision
result of a correlation direction.
For this reason, when a value of the temporary interpolation pixel
signal is a maximum value or a minimum value of a pixel signal of
the peripheral pixel, the correlation direction deciding unit 50
outputs a correlation direction by the second correlation detecting
unit 30 as a decision result. When a value of the temporary
interpolation pixel signal is not a maximum value or a minimum
value of a pixel signal of the peripheral pixel, the correlation
direction deciding unit 50 instructs the interpolation pixel
generating circuit 42 to output the temporary interpolation pixel
signal as an interpolation signal.
When the decision result of a correlation direction by the second
correlation detecting unit 30 is provided, the correlation
direction determining unit 41 outputs a pair of pixels based on the
decision result to the interpolation pixel generating circuit 42.
In such a case, the interpolation pixel generating circuit 42
generates a pixel signal of an interpolation pixel by using a pair
of pixels which has been newly set, and outputs as an interpolation
signal.
Next, operations of the embodiment configured as above will be
described referring to FIG. 7. FIG. 7 is an explanatory diagram
describing a selecting range of the peripheral pixel selecting
circuit 55. In FIG. 7, horizontal lines A, B, P and the respective
pixels represent the same lines and pixels as those of FIG. 2.
A correlation value of a pair of pixels of each correlation
direction candidate is delivered from the first and second
correlation detecting units 20 and 30 to the correlation direction
deciding unit 50. A correlation direction candidate of the first
correlation detecting unit 20 is based on a symmetric pair of
pixels, and a correlation direction candidate of the second
correlation detecting unit 30 is based on a symmetric pair of
pixels and a parallel pair of pixels.
The correlation direction deciding unit 50, first, selects one of
correlation direction candidates by the first correlation detecting
unit 20 to output it as a decision result of a correlation
direction to the correlation direction determining unit 41. The
correlation direction determining unit 41 outputs a pair of pixels
based on the inputted decision result to the interpolation pixel
generating circuit 42. The interpolation pixel generating circuit
42 generates a pixel signal by using the indicated pair of pixels,
and outputs as a temporary interpolation pixel signal to the
correlation direction deciding unit 50.
On the other hand, video signals of lines before and after an
interpolation line are also delivered to the peripheral pixel
selecting circuit 55. The peripheral pixel selecting circuit 55
selects peripheral pixels of an interpolation pixel. FIG. 7
illustrates a selecting range of the peripheral pixel selecting
circuit 55. FIG. 7 illustrates such an example that a peripheral
pixel selecting range of five pixels.times.two lines whose center
is an interpolation pixel is set. The peripheral pixel selecting
circuit 55 outputs pixel signals of peripheral pixels to the
correlation direction deciding unit 50.
The correlation direction deciding unit 50 compares a temporary
interpolation pixel signal with the pixel signals of peripheral
pixels to decide whether or not the temporary interpolation pixel
signal is a maximum value or a minimum value of the pixel signals
of peripheral pixels. When the temporary interpolation pixel signal
is not a maximum value or a minimum value of the pixel signals of
peripheral pixels, the correlation direction deciding unit 50
instructs the interpolation pixel generating circuit 42 to output
the temporary interpolation pixel signal as an interpolation
signal. Thereby, in such a case, the interpolation pixel generating
circuit 42 can obtain an interpolation pixel generated in high
accuracy by using a pair of pixels based on a correlation direction
by the first correlation detecting unit 20.
On the other hand, when the temporary interpolation pixel signal is
a maximum value or a minimum value of the pixel signals of
peripheral pixels, the correlation direction deciding unit 50
outputs a decision result of a correlation direction by the second
correlation detecting unit 30 to the correlation direction
determining unit 41. Thereby, the correlation direction determining
unit 41 indicates a pair of pixels based on the decision result to
the interpolation pixel generating circuit 42, and the
interpolation pixel generating circuit 42 generates a pixel signal
of an interpolation pixel by using the indicated symmetric pair of
pixels. Thus, in such a case, the interpolation pixel generating
circuit 42 generates an interpolation pixel by using a pair of
pixels based on such a correlation direction that wrong decisions
by the second correlation detecting unit 30 are few.
As described above, in the present embodiment, one of correlation
directions by the first and second correlation detecting units 20
and 30 is selected by comparing the pixel signals of peripheral
pixels with a temporary interpolation pixel signal generated by
using a pair of pixels based on a correlation direction by the
first correlation detecting unit 20, so that it is possible to
prevent wrong decisions and generate a highly accurate
interpolation pixel.
Meanwhile, according to whether or not the temporary interpolation
pixel signal is a maximum value or a minimum value of the pixel
signals of peripheral pixels, the correlation direction deciding
unit 50 decides whether or not there is a wrong decision in a
correlation direction by the first correlation detecting unit 20,
however, when a value of the temporary interpolation pixel signal
is parted by no less value than a prescribed threshold from values
of the pixel signals of peripheral pixels, the correlation
direction deciding unit 50 may decide that there is a wrong
decision.
Third Embodiment
FIG. 8 is a block diagram illustrating a third embodiment of the
present invention. In FIG. 8, the same code is attached to the same
component as that of FIG. 1, and the description will be
omitted.
The present embodiment also adopts the first and second correlation
detecting units 20 and 30. In the present embodiment, it is
determined to use which one of outputs from the first and second
correlation detecting units 20 and 30 by deciding the reliability
of a decision result of a correlation direction.
The present embodiment adopts a threshold setting circuit 65
instead of the horizontal high frequency detecting circuit 35, and
also adopts a correlation direction deciding unit 60 instead of the
correlation direction deciding unit 40. The threshold setting
circuit 65 sets a threshold for correlation values from the first
and second correlation detecting units 20 and 30. The threshold
setting circuit 65 provides the set threshold to the correlation
direction deciding unit 60. The threshold setting circuit 65 can
set a plurality of thresholds as a threshold. For example, the
threshold setting circuit 65 sets the first and second thresholds
which are mutually different.
When the reliability of a correlation direction is high which is
obtained according to correlation values from the first and second
correlation detecting units 20 and 30, it can be considered that
the correlation value for determining the correlation direction is
sufficiently small. When the value is relatively large even if it
is the minimum correlation value obtained by the first and second
correlation detecting units 20 and 30, it is possible that there is
a wrong decision on a correlation direction based on the minimum
correlation value.
Thus, in the present embodiment, when the minimum correlation value
from the first and second correlation detecting units 20 and 30 is
no less than a threshold set by the threshold setting circuit 65,
the correlation direction deciding unit 60 causes a decision result
of a correlation direction according to the correlation value to be
invalid.
For example, it is assumed that the relation of two thresholds set
by the threshold setting circuit 65 is such a relation that a first
threshold<a second threshold. In such a case, for example, when
the minimum correlation value from the first correlation detecting
unit 20 is no less than the set first threshold, the correlation
direction deciding unit 60 outputs a correlation direction by the
second correlation detecting unit 30 as a decision result, and also
when the minimum correlation value from the second correlation
detecting unit 30 is no less than the set second threshold, the
correlation direction deciding unit 60 outputs such a decision
result that a pair of pixels at a horizontally same position (upper
and under direction (vertical direction)) is set to an
interpolation pixel.
In addition, the threshold setting circuit 65 may output only one
threshold. In such a case, for example, when the minimum
correlation value from the first correlation detecting unit 20 is
no less than the set threshold, the correlation direction deciding
unit 60 may output a correlation direction by the second
correlation detecting unit 30 as a decision result, and on the
contrary, when the minimum correlation value from the second
correlation detecting unit 30 is no less than the set threshold,
the correlation direction deciding unit 60 may output a correlation
direction by the first correlation detecting unit 20 as a decision
result.
Meanwhile, when both of the minimum correlation values from the
first and second correlation detecting units 20 and 30 are no less
than the set threshold, the correlation direction deciding unit 60
may output a decision result indicating that a decision result of a
correlation direction by the first and second correlation detecting
units 20 and 30 is invalid. In such a case, the correlation
direction determining unit 41 may indicate a pair of pixels at a
horizontally same position (upper and under direction (vertical
direction)) of an interpolation pixel as pixels to be used for
generating an interpolation pixel.
When both of the minimum correlation values from the first and
second correlation detecting units 20 and 30 are smaller than the
set threshold, the correlation direction deciding unit 60 outputs a
correlation direction by the first correlation detecting unit 20 as
a decision result.
Next, the embodiment configured as above will be described.
A correlation value of a pair of pixels of each correlation
direction candidate is delivered from the first and second
correlation detecting units 20 and 30 to the correlation direction
deciding unit 60. The correlation direction deciding unit 60 is
provided with a threshold which is set for a minimum correlation
value from the threshold setting circuit 65. It is assumed that two
thresholds whose relation is a relation: a first threshold<a
second threshold is outputted from the threshold setting circuit
65.
The correlation direction deciding unit 60 decides whether or not a
minimum correlation value by the first and second correlation
detecting units 20 and 30 is no less than the first threshold. When
the minimum correlation value is smaller than the first threshold,
the correlation direction deciding unit 60 outputs a correlation
direction by the first correlation detecting unit 20 as a decision
result. Thereby, the interpolation pixel generating circuit 42
generates an interpolation pixel in high accuracy by using a pair
of pixels based on a correlation direction by the first correlation
detecting unit 20.
Among correlation values from the first and second correlation
detecting units 20 and 30, when a minimum correlation value from
the first correlation detecting unit 20 is no less than the first
threshold and smaller than the second threshold, the correlation
direction deciding unit 60 outputs a decision result of a
correlation direction by the second correlation detecting unit 30
to the correlation direction determining unit 41. Thereby, the
correlation direction determining unit 41 indicates a pair of
pixels based on the decision result to the interpolation pixel
generating circuit 42, and the interpolation pixel generating
circuit 42 generates a pixel signal of an interpolation pixel by
using the indicated symmetric pair of pixels. Thereby, in such a
case, the interpolation pixel generating circuit 42 generates an
interpolation pixel by using a pair of pixels based on a
correlation direction in which wrong decisions by the second
correlation detecting unit 30 are few.
When both of minimum correlation values from the first and second
correlation detecting units 20 and 30 are no less than the second
threshold, the correlation direction deciding unit 60 outputs a
decision result indicating that correlation directions by the first
and second correlation detecting units 20 and 30 are invalid to the
correlation direction determining unit 41. In such a case, for
example, the interpolation pixel generating circuit 42 generates an
interpolation signal by using a symmetric pair of pixels at a
horizontally same position as an interpolation pixel.
As described above, in the present embodiment, when the minimum
correlation value from the first and second correlation detecting
units 20 and 30 is no less than the threshold, it is decided that
the correlation direction by the first and second correlation
detecting units 20 and 30 is invalid. Thereby, it is possible to
prevent a wrong decision, and also generate an interpolation pixel
in high accuracy.
Fourth Embodiment
FIG. 9 is a block diagram illustrating a fourth embodiment
according to the present invention. In FIG. 9, the same code is
attached to the same component as that of FIG. 1, FIG. 6, and FIG.
8, and the description will be omitted.
The present embodiment is a combination of the first to third
embodiments. That is, the present embodiment includes the
horizontal high frequency detecting circuit 35, the peripheral
pixel selecting circuit 55 and the threshold setting circuit 65. A
correlation direction deciding unit 70 is provided with correlation
values from the first and second correlation detecting units 20 and
30, and decides a correlation direction by using outputs of the
horizontal high frequency detecting circuit 35, the peripheral
pixel selecting circuit 55 and the threshold setting circuit
65.
Next, operations of the embodiment configured as above will be
described referring to FIG. 10. FIG. 10 is a flowchart illustrating
an example of a decision processing of a correlation direction by
the correlation direction deciding unit 70.
A correlation value of a pair of pixels of each correlation
direction candidate is delivered from the first and second
correlation detecting units 20 and 30 to the correlation direction
deciding unit 70. In addition, a decision result of the horizontal
high frequency detecting circuit 35, a pixel signal from the
peripheral pixel selecting circuit 55, a threshold from the
threshold setting circuit 65, and a temporary interpolation pixel
signal from the interpolation pixel generating circuit 42 are also
inputted to the correlation direction deciding unit 70.
The correlation direction deciding unit 70 decides a correlation
direction based on at least one of the decision result of a
horizontal high frequency component, a pixel signal of a peripheral
pixel, a threshold for a minimum correlation value, and a temporary
interpolation pixel signal. Description will be made assuming that
two thresholds whose relation is a relation: a first threshold<a
second threshold are provided from the threshold setting circuit 65
to the correlation direction deciding unit 70.
When a minimum value of correlation values from the first and
second correlation detecting units 20 and 30 is no less than the
second threshold, the processing proceeds from any one of step S1
to S3 of FIG. 10 to step S4. In such a case, the correlation
direction deciding unit 70 decides that the reliability of any
correlation decision based on the first and second correlation
detecting units 20 and 30 is low, and at step S5, outputs a
decision result of a correlation direction for generating an
interpolation pixel based on a symmetric pair of pixels at a
horizontally same position (upper and under direction (vertical
direction)).
When a minimum value of correlation values from the first and
second correlation detecting units 20 and 30 is smaller than the
second threshold, the decision result of a correlation direction is
obtained at step S1 to S3. That is, at step S1, the correlation
direction deciding unit 70 decides whether or not a pair of pixels
of a correlation direction candidate is included in a horizontal
high frequency component area. When the pair of pixels of a
correlation direction candidate is included in the horizontal high
frequency component area, the correlation direction deciding unit
70 proceeds from step S1 to step S6 through step S4, and outputs a
correlation direction by the second correlation detecting unit 30
as a decision result. That is, in such a case, it is possible to
generate an interpolation pixel without wrong decisions even in a
fine pattern part.
When the pair of pixels of a correlation direction candidate is not
included in the horizontal high frequency component, at step S2,
the correlation direction deciding unit 70 decides whether or not a
temporary interpolation pixel signal is a maximum value or a
minimum value of pixel signals of peripheral pixels. When the
temporary interpolation pixel signal is a maximum value or a
minimum value of pixel signals of peripheral pixels, the processing
proceeds from step S1 to step S6 through S4, and a correlation
direction by the second correlation detecting unit 30 is outputted
as a decision result. That is, in such a case, the reliability of
the temporary interpolation pixel signal based on the first
correlation detecting unit 20 is low, and the reliability of
correlation values from the second correlation detecting unit 30 is
relatively high, so that a highly reliable interpolation pixel is
generated by generating an interpolation pixel based on the second
correlation detecting unit 30.
At step S2, when deciding that the temporary interpolation pixel
signal is not a maximum value or a minimum value of pixel signals
of peripheral pixels, the correlation direction deciding unit 70
proceeds to next step S3, and decides whether or not a correlation
value from the first correlation detecting unit 20 is smaller than
the first threshold. When the correlation value from the first
correlation detecting unit 20 is no less than the first threshold,
the correlation direction deciding unit 70 decides the processing
to proceed from step S1 to step S6 through step S4, and outputs a
correlation direction by the second correlation detecting unit 30
as a decision result.
At step S3, when it is decided that the correlation value from the
first correlation detecting unit 20 is smaller than the first
threshold, at next step S7, a correlation direction by the first
correlation detecting unit 20 is outputted as a decision result.
That is, in such a case, it is possible to generate an
interpolation pixel in high accuracy.
As described above, in the present embodiment, the same effect as
each of the above embodiments can also be obtained.
The above embodiments have such an effect that interpolations can
be executed without inducing breaks, unsharpness, jaggy, etc. even
in a thin line-like pattern part and the like.
Having described the embodiments of the invention referring to the
accompanying drawings, it should be understood that the present
invention is not limited to those precise embodiments and various
changes and modifications thereof could be made by one skilled in
the art without departing from the spirit or scope of the invention
as defined in the appended claims.
* * * * *